This proposal sets the stage for clinical trials of gene therapy for hemophilia A. Here we concentrate on the use of adeno-associated viral (AAV) vectors with new capsid serotypes for liver directed gene therapy in a large animal model, namely, the hemophilia A dog. This work builds upon our previous success in "curing" the hemophilia A mouse using FVIII gene delivery in two new AAV serotypes, AAV8 and AAV9. Our studies in hemophilia A dogs, while more modestly successful, have also been encouraging with both new AAV serotypes. In this application, we evaluate further these AAV serotypes, beginning with delivery of FVIII cDNA in heavy chain and light chain vectors. We test responses at two different doses for both AAV8 and AAV9, and compare intraportal to intravenous routes of vector administration, We then go on to develop an effective, smaller single chain vector that can be produced in high titers, and test this single chain vector using the best combination of serotype, dose, and route of administration. We will also carry out therapy with our best conditions in young dogs, in an effort to initiate treatment before the onset of serious morbidity and also to prevent inhibitor formation to rFVIII used for conventional therapy. Lastly, we will determine whether FVIII can be readministered to dogs using a different serotype vector from the vector used in the original virus infusion. In all of these studies significant efforts will be made to test the safety of the new AAV vectors, concentrating on studies of neutralizing antibodies, development of transaminitis, and germline transmission of virus. At the end of these studies, we expect to have discovered treatment conditions that consistently produce long-term therapeutic levels of FVIII (>10% of normal) in a large animal, the hemophilia A dog. These results should make clinical trials with AAV vectors an appropriate next step.